pub fn tokenize_as_ids(
text: &str,
dictionary: &TokenDictionary,
) -> Vec<QueryToken>Expand description
Tokenizes text and returns QueryTokens directly, avoiding string allocation.
This is the primary tokenization function for query processing. Tokens are classified against the dictionary immediately:
- Known tokens โ QueryToken::Known(KnownToken)
- Unknown tokens โ QueryToken::Unknown
- Stopwords โ QueryToken::Stopword
ยงReturns
A vector of QueryTokens (no string allocation).